What's intriguing about the equatorial behaviors of ENSO and QBO is that they should be driven by congruent forcing factors. ENSO is much more complex so it has a few factors to choose from, but the QBO is driven primarily by the Draconic cycle (with a small seasonal factor).

After modeling ENSO to a rather precise degree, the Draconic forcing was used as a starting point to model QBO. It was then allowed to vary slightly to enable a gradient-search fit.

This is how the two forcing factors align:

![cc](https://imageshack.com/a/img922/8009/7EyJNs.gif)

This is how well the QBO model fits the 30 hPa QBO data using essentially the same Draconic profile as was used for ENSO:

![qbo](https://imageshack.com/a/img922/6417/tDYJxZ.gif)

The odds of that happening by chance is exceedingly low. When fitting the ENSO data, there was no training against QBO, yet very close to the same factor worked for QBO. That's a type of cross-validation that's vitally important for climate modeling.